CN104823421B

CN104823421B - Method and corresponding transmission equipment for multiple-carrier signal transmission

Info

Publication number: CN104823421B
Application number: CN201380062817.5A
Authority: CN
Inventors: M.拉诺伊塞勒; B.杰汉
Original assignee: France Telecom SA
Current assignee: Orange SA
Priority date: 2012-10-05
Filing date: 2013-09-30
Publication date: 2018-01-23
Anticipated expiration: 2033-09-30
Also published as: EP2904754B1; WO2014053757A1; US20150271001A1; US9455858B2; EP2904754A1; CN104823421A

Abstract

The present invention relates to the method for transmitting ofdm signal, wherein pre-transmission processing includes：The data for representing source signal are mapped on the complex symbol Xn for belonging to constellation, 0≤n<M, n and M are integers；M symbol Xn is transformed into M and has corrected symbol X ' n so that X ' n=Xn+dn, wherein dn are plural amendments；On M that M corrected to symbol X ' n are mapped in N number of carrier wave of OFDM modulators, to generate ofdm signal, N is integer, M≤N so that the conversion includes accumulator initialization step, and for each n ranks carrier wave (n is from 1 to M 1), comprising：In accumulator, respectively the J sample corresponding J time samples for the n rank carrier waves with being mapped by X ' n of being added up on already present J sample the step of (1051,1054)；Detect peak value in the J sample obtained from the accumulation step of the ranks of n 1, and by the peak value compared with the corresponding time samples in J time samples of n rank carrier waves the step of (1052), and transmit plural Correction and Control item of information (Poln)；Symbol X ' n amendment step (1053) has been corrected to obtain to determine plural correction value dn according to plural Correction and Control item of information (Poln).

Description

Method for multicarrier signal transmission, and corresponding transmission device

1. Field of the invention

The field of the invention is radio frequency communication, in which multi-carrier modulation, in particular of the OFDM (orthogonal frequency division multiplexing) type, is used.

More precisely, OFDM modulation is increasingly used for digital transmission, in particular for transmission channels with multipath. The multi-carrier modulation technique makes it possible to significantly remove the interference between symbols that is typically observed when using single-carrier modulation for a multi-path channel. Furthermore, the technique provides very good spectral efficiency and makes it possible to save radio frequency spectrum resources through the implementation of a single frequency network.

Due to the robustness inherent to multipath channels and frequency selective channels, OFDM modulation is used in particular, but not exclusively, for wireless (WiFi) local area networks, 3GPP LTE ("3 rd generation partnership project" and "long term evolution") cellular mobile radio telephones, or even ADSL ("asymmetric digital subscriber line"), but also for standards such as digital TV broadcasting involving Digital Audio Broadcasting (DAB), including in particular DVB-T (terrestrial digital video broadcasting) or even the new DVB-T2 standard.

2. Background of the invention

2.1 disadvantages of OFDM modulation

The main drawback of OFDM technology is inherent to strong amplitude fluctuations of the modulated signal envelope and thus significant variations of the instantaneous power.

In practice, in the time domain, these independently modulated multi-carriers sum in terms of power most of the time, but are also coherent, which may produce instantaneous power peaks that exceed the average power of the signal by more than 10 decibels at some time instants.

The peak-to-average ratio (PAPR), a factor that characterizes the power peak level relative to the average power of the signal, of the transmitted signal is therefore generally very high and increases with the number of carriers N.

Power amplifiers have non-linear characteristics that, together with amplification with so-called strong PAPR signals, will introduce distortion: side lobe spectrum is raised, harmonic waves are generated, interference is generated between nonlinear code elements, and interference is generated between carriers. Therefore, these distortions significantly cause transmission errors and degraded Bit Error Rates (BER).

2.2 definition of PAPR

More precisely, according toIn a particular embodiment, an OFDM signal of band B is used consisting of a total of N regularly modulated orthogonal carriers, which are separated by a frequency interval Δ f such that B = N. For a given OFDM block, each carrier is attributed to symbol X of the constellation (QPSK, MAQ16, etc.) _n And (5) modulating. An inverse fourier transform (IFFT) of the frequency signal of band B then provides the signal x (t) to be transmitted in the time domain. In the time domain, the duration of an OFDM block is n.te =1/Δ f, where Te is the sampling period, expressed as follows:

suppose variable X _n Is random, statistically independent and focused, from which the PAPR of the OFDM signal is deduced, expressed as:

it is observed that with this definition of PAPR, and with an IFFT where x (t) is a discrete random variable, in a particular but also very rare case, the PAPR can become as large as N, where

In practice, the peak PAPR of a given amplitude occurs with a certain probability. It is obviously unlikely that the signal amplitude is as large as N, especially as N becomes larger. Also, in general, to characterize the PAPR of an OFDM system, a Complementary Cumulative Distribution Function (CCDF) is used, which provides the probability that the signal amplitude exceeds a certain threshold. This function is the most commonly used function to characterize PAPR reduction systems, expressed as follows:

CCDF _PAPR ＝Pr[PAPR(X _L )>r,L＝1]

≈1-(1-e ^-r ) ^N

in practice, the equation indicates, for example, that if the digital/analog and/or analog/digital converters and the power amplifiers are not operated with a dynamic difference of at least 12.2 db between the average power and the peak power (for an amplifier this represents a 1 to 16 operating power ratio), the signal will not be transmitted correctly when the signal contains 2048 carriers, e.g. without a sample saturation of at least one of the 100 symbols.

Below this limit the signal will be clipped, or at least severely distorted with the transmission residual and the conditions of reception.

2.3 Prior Art for PAPR reduction

In the literature, several techniques have been proposed to alleviate this problem.

One current solution is to ensure that the operating range of the amplifier remains limited to the linear amplification interval, which unfortunately limits the range of the amplifier (a few% rather than the usual 50%) and thus leads to a significant increase in transmitter consumption. This is a strong constraint for the use of OFDM, especially in mobile terminals, bearing in mind that the consumption of the power amplifier may represent more than 50% of the total consumption of the terminal.

The second approach is to apply a constraint or coding to the transmitted data sequence to limit PAPR. The method consists in constructing a set of code words that minimize the PAPR. A number of techniques have been proposed for constructing such codes. The advantage of this solution is that it does not introduce any distortion. On the other hand, the disadvantage of spectral efficiency is that it does not provide any coding gain. Furthermore, so far, due to great computational complexity, its application range is limited to OFDM modulators with a small number of carriers N.

A third approach, commonly referred to as the "TI-CES (zhuyin-constellation extension scheme) technique", proposes to increase the number of points of the constellation modulating the OFDM carrier, so that for the points of the original constellation there can be several possible corresponding coordinates in the new constellation. According to this approach, this added degree of freedom is used to generate signals with lower PAPR. However, this approach has several disadvantages, since the added symbols have higher power levels, constellation expansion will bring about an increase in the average power of the signal. Furthermore, the selection of the best possible coordinate for each point requires an increase in the complexity of the calculations implemented, making it unsuitable for hardware implementations of real-time signal processing.

The fourth approach, commonly referred to as the "CD (constellation distortion) technique", is also based on constellation modification and, under the assumption that the output level of the transmission amplification is defined by the strongest PAPR peaks, and if the amplitude of these peaks can be reduced, the transmitted power will increase. According to this technique, for a given distortion rate, an optimization problem known as convex is solved to generate an OFDM signal with a minimum overall PAPR level. However, this method requires a large increase in average output power to neutralize the loss in terms of signal-to-noise ratio. Furthermore, the computational complexity implemented increases exponentially as the constellation order becomes higher.

The fifth approach, commonly referred to as "ACE (active constellation extension) technique", is also based on constellation modification and depends on the displacement achieved in a direction away from the decision axis. However, in the same way as the two previous methods, this technique is characterized by a low efficiency for high order constellations due to the increased average power of the signal and the very high computational complexity.

A sixth approach, commonly referred to as the "TR (tone reservation) technique", proposes to reserve some of the carriers of the OFDM multiplex, which do not carry information, but carry transmission optimized symbols to reduce PAPR. The optimization of these symbols can be achieved by using for example a convex optimization algorithm of the SOCP (second order cone programming) type. Just like the previous method, this solution does not add any distortion to the transmitted signal, but the greatest drawback of this method is that a certain number of carriers must be reserved to be able to reduce the PAPR significantly. These carriers cannot be used to transmit useful information data, which leads to a reduction in spectral efficiency.

A seventh approach, called "selective mapping", consists in applying a phase rotation to each symbol of the sequence to be transmitted. Multiple phase rotation modes may be defined. For each mode applied to the sequence to be transmitted, an operation is carried out to obtain a corresponding OFDM signal, this signal showing the lowest PAPR being transmitted. Also, this technique does not add any distortion, but it requires a very high reliability in communicating to the receiver the rotation sequence used in the transmission, which brings about a reduction in spectral efficiency and a significant increase in system complexity to transmit the applied rotation pattern via a dedicated channel. Furthermore, if the transmission is erroneous, the entire OFDM frame will be lost. This also increases the complexity of the transmission, since multiple processing operations must be implemented in parallel to select the most efficient. Other implemented processing operations are meaningless and not used.

The last approach is a "clipping" (or limiting) technique, which consists in clipping the amplitude of a signal when it exceeds a predetermined threshold. However, this clipping is nonlinear in nature and introduces distortion of the transmitted signal, which is reflected not only in degraded BER but also in the rise of PSD (power spectral density) sidelobes.

In this particular context, the inventors have then recognized that new techniques are needed to make it possible to improve the reduction of PAPR and to keep the implementation simple.

3. Summary of the invention

The present invention proposes a new solution in the form of a method for transmitting OFDM signals, which does not show all these drawbacks of the prior art, the processing of OFDM before transmission comprising:

-mapping data representing the source signal onto complex symbols Xn belonging to a constellation, 0 ≦ n < M, n and M being integers,

-transforming the M symbols Xn into M modified symbols X 'n, such that X' n = Xn + dn, where dn is a complex modification,

mapping the M modified symbols X' N onto N carriers of an OFDM modulator to generate an OFDM signal, N being an integer, M ≦ N,

the transformation includes:

-an initialization step of the accumulator,

and for each nth order carrier (n from 1 to M-1) comprises:

-a step of accumulating, in an accumulator, J time samples corresponding to J samples of the order n carriers mapped by X' n, J being an integer, with J samples already present, respectively,

-a step of detecting a peak value among the J samples resulting from the step of accumulating to the order n-1 and comparing this peak value with a simultaneous time sample among the J time samples of the carrier of order n and delivering a complex correction control information item,

-a correction step of determining a complex correction value dn as a function of the complex correction control information item to obtain a corrected symbol X' n.

The invention also relates to a device for transmitting OFDM signals. According to the invention, the transmission device comprises:

-a mapping module for mapping data representative of the source signal onto complex symbols Xn belonging to a constellation, 0 ≦ n < M, n and M being integers,

-a transformation module for transforming the M symbols Xn into M modified symbols X 'n such that X' n = Xn + dn, where dn is a complex modification,

-an OFDM modulator with N carriers to generate an OFDM signal from M modified symbols X' N mapped on M carriers, M being an integer, M ≦ N, and when M ≠ N, N-M carriers are unmapped.

The transformation module includes:

-a construction and accumulation module for accumulating J time samples corresponding to J samples of the current carrier of order n mapped by X' n to the existing J samples, 1 ≦ J,

-means for detecting the peak of the J samples at the output of the building and accumulation module of order n-1 and comparing this peak with the simultaneous time sample of the J time samples of the current carrier of order n and delivering an item of complex modification control information,

-a correction module for determining a complex correction value dn as a function of the complex correction control information item to obtain a corrected symbol X' n.

Such a transmission device is particularly suitable for implementing the transmission method according to the invention.

The present invention thus relies on a new and inventive approach to reduce PAPR of OFDM signals.

Rather, the present invention makes it possible to improve the performance level of PAPR reduction with lower computational complexity compared to the prior art.

Furthermore, the present invention provides great flexibility in constellation modification compared to that carried out by TI-CES, CD, ACE and TR technologies.

In fact, the method according to the invention modifies in a continuous and controlled manner, in the frequency domain, the constellation symbols modulating the carriers of the OFDM block, before implementing the frequency-time transformation.

According to a specific embodiment, the frequency-time transform is a discrete inverse fourier transform. According to a specific embodiment, the inverse fourier transform is a fast fourier transform (IFFT). In the following, the acronym IFFT refers to a frequency-time transformation carried out by an OFDM modulator, which makes it possible to generate an OFDM signal.

To reduce PAPR, the present invention uses modified real-time synchronization of a carrier, called the current carrier, which is associated with a previously modified carrier of the same OFDM block. "correction of the carrier n" is understood to mean the correction of the symbol Xn by the value dn to produce the symbol X' n, which maps the carrier n in the OFDM modulation implemented by the OFDM modulator.

The synchronization implements, among other things, detection of PAPR peaks in a set of time samples representing a sum of time responses of previously modified carriers based on constellation symbols modulating the current carrier for modification.

Then, a complex correction control information item is obtained by considering time samples of the carrier to be corrected which occur simultaneously with the peak value which has been detected as required. The complex correction control information item is then used to determine the correction to be made to the complex coordinates of the constellation symbol modulating the current carrier to be corrected.

Thus, after accumulation is complete, the above detection and correction steps are carried out for each of the M mapped carriers, and the method implements in the frequency domain (i.e. before the frequency-time transformation) a "pre-construction" of the corrected time signal, which consists of J samples and is associated with an OFDM block of N carriers mapped by M (M ≦ N) corresponding symbols.

To this end, the transform is constructed in parallel with X' _n-1 J time samples corresponding to J samples of the mapped n-1 order carrier are accumulated in an accumulator. The accumulation means that the J samples are added to the J samples already present in the accumulator, respectively. After each accumulation, the transform detects the occurrence of a PAPR peak in the accumulated J samples. Moreover, the transform constructs J time samples corresponding to the current carrier of order n in parallel. The transform compares the detected PAPR peak among the J samples resulting from the accumulation to order n-1 with the time samples of the J time samples of the current carrier of order n, located at the same index as in the block of size J. The transformation determines a complex correction control information item, making it possible to determine a complex correction value. In particular, if the peak and time samples have respective amplitudes and phases, which will result in amplification of the peak after accumulation to n-th order, complex correction values are determined to counter the amplification effect.

According to particular implementation features, the method develops the calculation of an inverse discrete Fourier transform of a block of N carriers, where M carriers are coded by a symbol X 'by implementing the calculation carrier by carrier' _n Mapping, for a given carrier n, by X 'by calculating each time sample I (I varies between 0 and J-1)' _n And (6) mapping. The method thus determines X' _n Time samples associated with the mapped carriers. The accumulation consists in adding J time samples of the current carrier n to the first J accumulated samples associated with the carrier up to the order n-1, respectively.

It should be noted that the frequency-time transform, typically an inverse fourier transform, may be applied to the M carriers mapped by the M symbols under consideration, less than the number of carriers N of the OFDM modulator. In practice, it is common practice to reserve carriers to insert preambles therein, limit the problem of spectral overlap by setting the edge carriers to 0, and avoid the continuous portion by setting the center carrier to 0.

It should be noted that even in the frequency domain, the term "pre-constructed" means "capable" of determining time samples of the signal response obtained after the IFFT is determined. In fact, the object of the invention is to modify the complex coordinates of the constellation symbols modulating the current carrier in the frequency domain.

The term "pre-constructed" is therefore associated with time samples that may be modified prior to implementation of the OFDM modulator.

A "complex number" is to be understood to mean "may have a real and/or imaginary value, such that this value is defined as v = a + bj", for example.

This approach thus relies on the fact that each carrier of an OFDM block can be modified to result in an overall modification of the OFDM time signal.

The modification of the time signal is optimized by virtue of the fact that the complex displacement of the constellation coordinates is determined as a function of the complex modification control information item. As described below, this dependency between the complex displacement of the constellation coordinates and the complex modification control information item makes it possible to construct a new constellation. The new constellation may correspond, for example, in a particular manner to a modified constellation or to a combination of constellations from the aforementioned constellation modification techniques (i.e., TI-CES, CD, ACE, and TR techniques).

In this configuration, one advantage of the technique proposed according to the invention is thus that it is possible to increase the efficiency by making it possible to associate a plurality of constellation modification techniques, the various drawbacks of which, respectively, can be neutralized to some extent, by means of a correction control based on the implementation of a real-time synchronization between the current carrier to be corrected and the carrier that precedes it in time and has been previously corrected.

According to a particular aspect of the invention, the step of detecting J samples resulting from the step of accumulating to order n is carried out in association with the step of generating J time samples associated with the current carrier with index n with a delay corresponding to a predetermined length of the clock cycle.

The implementation of said delay according to a predetermined clock cycle makes it possible in particular to obtain synchronously a complex correction control information item in order for it to be transmitted at the instant at which the step of correcting the constellation symbol modulating the current carrier of order n to be corrected occurs, making it possible to obtain a constellation symbol of order n +1X 'occurring in the step of accumulating' _n 。

Preferably, the detected peak is a power peak, the J time samples of the nth order carrier are complex, and the complex modification control information item belongs to at least one of the following categories:

-a complex modified control information item whose real (respectively imaginary) part is positive when the peak value and the sign of the real (respectively imaginary) part of the corresponding complex time sample are the same.

-a complex modified control information item whose real (respectively imaginary) part is negative when the peak value and the sign of the real (respectively imaginary) part of the corresponding complex time sample are opposite.

A complex modification control information item, the real (respectively imaginary) part of which is 0, when the power amplitude of the time sample of the corresponding detected peak is less than a predetermined threshold, or when said current carrier with index n is said to be "reserved" without modification thereof (for example in the case where the insertion of preamble symbols occurs before the implementation of the method according to the invention).

Thus, the complex modification control information item makes it possible to control reduction of PAPR by considering symbol correlation between a peak value representing a previously modified and accumulated pre-constructed time sample of a carrier and a complex time sample of a current carrier to be modified occurring simultaneously with the detected peak value.

It is thus possible to control the correction carried out from one carrier to another in an optimized manner by taking into account the correlation between the current carrier to be corrected and the peaks detected from the accumulation of the previously corrected carriers, or by taking into account the identification of the so-called "reserved" carrier whose constellation symbols must remain unchanged. According to particular aspects of the present invention, it is thus possible to not modify a predetermined or previously identified "reserved" carrier.

Advantageously, the correction step implements a summation of the coordinates of the constellation symbol and coordinates representing complex displacements of the constellation symbol on the X and Y axes of a complex plane of the constellation of the symbol, said complex displacements being selected by means of complex correction control information items from complex displacements belonging to at least one of the following categories:

-a real (respectively imaginary) displacement of the complex displacement, which is negative when the complex modification control information item contains a positive real (respectively imaginary) part;

-a real (respectively imaginary) displacement of the complex displacement, which is positive when the complex modification control information item contains a negative real (respectively imaginary) part;

the real (respectively imaginary) displacement of the complex displacement, which is 0 when the complex modification control information item contains a real (respectively imaginary) part of 0.

"real displacement of a complex displacement" is understood to mean a displacement along the real axis of the real part of the complex displacement. "imaginary displacement of a complex displacement" is understood to mean a displacement along the imaginary axis of the imaginary part of the complex displacement. In practice, the displacements of the real and imaginary parts of the constellation symbols are independent of each other. For example, the real part of the constellation symbol may be positively shifted with respect to the real axis, while the imaginary part may be negatively shifted with respect to the imaginary axis.

The invention therefore proposes to control the complex displacement of each constellation symbol of a modulated carrier on the X and Y axes of the complex plane of the constellation of the symbol. A controlled complex displacement of the constellation symbols is thus obtained, which may distinguish one constellation symbol modulating a carrier from another constellation symbol modulating another carrier.

In other words, it is for example possible to modify the real and/or imaginary part of the constellation symbols of a carrier with a modulation index N + g (g being an integer such that 0 ≦ N + g ≦ N-1) according to a complex displacement having a positive real (respectively imaginary) part, while modifying the real (respectively imaginary) part of the constellation symbols of a carrier with a modulation index N according to a complex displacement having a negative real (respectively imaginary) part.

Since the nature of the complex shifts is controlled, for example when a CD type of mapping modification is selected, the complex shifts implemented may result in constellation points being retained in their decision regions or original constellations, or when an ACE or CES type of modification is selected according to another exemplary modification, they are moved outside.

Furthermore, a 0 complex shift results in the original constellation point being retained, i.e. not modified.

According to a particular aspect of the invention, the absolute value of the real (respectively imaginary) part of the complex displacement is determined from one carrier to the other of the OFDM block, which corresponds to a predetermined value. For example, the complex shift value is equal to 0.25 or 0.5 times the distance from the constellation point to the decision region boundary.

The stability of the absolute value of the real (respectively imaginary) part of the complex displacement makes it possible, among other things, to limit the complexity of the implementation of the invention.

According to an embodiment, J = l.n time samples are constructed according to the method of the invention, where L is an integer greater than or equal to 1. When L =1, J = N. When L >1, for example when L =2 or 4, is oversampling. Such oversampling advantageously makes it possible to obtain greater resolution in PAPR reduction.

According to a particular embodiment, the transmission method successively carries out at least the following sub-steps:

-applying an inverse fast Fourier transform to the real and imaginary parts of the constellation symbols, centered around a frequency equal to Fe/2,

-transposing the real and imaginary parts to baseband,

oversampling the real and imaginary parts in the baseband by a factor L (e.g. with L = 2) at a frequency equal to 2.Fe,

-low-pass filtering of the real and imaginary parts,

-modulating said portion at a carrier frequency of Fe/2.

More specifically, the method according to the invention implements the following equation:

wherein:

‐A _n and B _n Is to the index nThe real and imaginary components of the constellation symbol on which the current carrier is modulated,

-S is all pre-constructed real time samples associated with an OFDM block,

‐

-and 0 ≦ l < J = l.n =2.n.

According to an embodiment of the device, the detection module detects a maximum power peak.

According to an embodiment of the apparatus, the transformation module further comprises:

-means for generating J time samples associated with a current carrier with index n, 0 ≦ n < M,

the build and accumulate module comprises:

-means for constructing J time samples of the current carrier of order n mapped by X' n, and

-an accumulation module.

The invention also relates to a computer program comprising instructions for implementing the transmission method described above when the program is executed by a processor.

4. Description of the drawings

Other features and advantages of the invention will appear more clearly on reading the description of specific embodiments which follows, given purely by way of illustrative and non-limiting example, and the accompanying drawings, in which:

figures 1A and 1B show simplified block diagrams of a processing architecture for OFDM signals and a PAPR reduction system according to the invention, respectively;

fig. 2 illustrates the main steps of a transmission method according to an embodiment of the invention;

FIG. 3 illustrates a detailed block diagram of a PAPR reduction system according to an embodiment of the present invention;

FIGS. 4A and 4B illustrate different calculation sub-steps carried out by the method according to the invention;

figures 5A to 5C illustrate different examples of modified constellations obtained according to the invention;

fig. 6 illustrates the structure of the transmission device according to the invention.

5. Detailed description of the preferred embodiments

5.1 general principles

The present invention relies on the use of control over the modification of the constellation of the modulated OFDM signal to optimally reduce the peak-to-average ratio, or PAPR.

More precisely, the invention implements a "pre-construction" of a modified real digital signal representative of the signal obtained at the output of a transmission device with reduced PAPR.

The method according to the invention detects PAPR peaks on a carrier-by-carrier basis during "pre-construction" of the digital signal. When these peaks occur, the method according to the invention transmits a complex correction control information item which makes it possible to optimize the modification of the constellation associated with the carriers of the OFDM signal in order to reduce these peaks.

In particular, the control information item is obtained by taking into account both the current carrier to be modified and the previously modified carrier. The invention thus makes it possible to adapt the constellation of the signal to be transmitted carrier by carrier.

A new signal modulation constellation allowing PAPR reduction is thus obtained according to the present invention.

An exemplary architecture for processing an OFDM signal to reduce PAPR will be described below in conjunction with fig. 1A. According to the embodiment illustrated in fig. 1A, the OFDM signal is processed according to a series of steps:

for transmission through the transmitter 1000:

-generating 101 source data;

-encoding and interleaving 102 said data and transmitting interleaved data;

-modulating 103 said interleaved data, for example according to QAM modulation, consisting in mapping the interleaved data representative of said source signal on complex symbols Xn belonging to a constellation;

-inserting 104 pilots;

-modifying 105 the symbols of the OFDM block to reduce PAPR according to the method of the invention;

OFDM modulation 106, for example implementing an Inverse Fast Fourier Transform (IFFT), and transmitting OFDM symbols, the modulation consisting in mapping the modified symbols on the carriers of an OFDM modulator to generate an OFDM signal;

-transmitting 107 said OFDM signal over a transmission channel 108, the transmission typically being accompanied by noise, such as white gaussian noise 109;

upon reception by the receiver 1010:

-receiving 110 a so-called reception signal;

-according to a particular embodiment, OFDM demodulating 111 said received signal by implementing a Fast Fourier Transform (FFT) and transmitting the transformed received signal;

-a channel estimation 112;

-demodulating 113 said transformed received signal and transmitting a demodulated signal;

-deinterleaving 114 and decoding 115 said demodulated signal;

-determining a bit error rate.

The invention thus proposes a specific correction technique 105 which makes it possible to reduce the PAPR efficiently and which is easy to implement. Furthermore, the correction according to the invention is only implemented in the transmission and does not require any modification of the existing receiver.

According to the illustration of fig. 1A, the transmission method according to the invention is carried out after the insertion 104 of the pilot. According to other embodiments, the insertion can be performed in an interleaved manner or immediately following the steps of the method according to the invention. For simplicity of description, it is considered that the method according to the invention considers a block of N symbols, whose size is equal to the total number N of carriers of the OFDM modulator (thus M = N). The block of M symbols may also be smaller than N to account for reserved carriers whose values are set elsewhere, or unmodulated carriers at the edges of the spectrum.

The PAPR reduction method 105 according to the present invention will be described in terms of the block diagram of fig. 1B. More precisely, the main steps of the transmission method according to the invention are implemented in the frequency domain between the modulation 103 and the conventional steps of OFDM modulation 106, for example implementing an inverse fast fourier transform IFFT.

More precisely, according to fig. 1B, the method according to the invention corresponds to a synchronous system of the retrospective type ("feedback" type).

Specifically, the method runs in real-time at a frequency Fe (e.g., the sampling frequency of the source data) according to the clock timing. Furthermore, the method is non-iterative, i.e. the corrections of the block involving N carriers (N equally corresponding to the size of the fourier transform and the inverse fourier transform) are all calculated at the frequency Fe over the duration of N samples.

The method consists in "pre-constructing" which, before the Inverse Fast Fourier Transform (IFFT) 106, uses the building and accumulation module 1051, the detection and comparison module 1052 and the modification module 1053 represented in fig. 1B, a real time signal can be obtained at the output of the transmission device from a series of different carriers of a block of OFDM signals modulated by a constellation symbol.

For each carrier of the OFDM modulator modulated by symbol Xn, if this carrier is transformed separately in the time domain, the building and accumulation module 1051 calculates simultaneously all J samples of the time response of this carrier mapped by symbol X' n, which can be obtained after the inverse fourier transform carried out by the OFDM modulator. The different time responses are then accumulated by the build and accumulate module 1051, carrier by carrier, up to the current carrier of order N-1.

In order n, the signal vector S' n-1 at the output of the building and summation module]From the corrected up to the order of n-1 (n)&gt, 1) carrier sum of ₀ The time sample composition of the mapped uncorrected 0 th order carrier computed in parallel. At the end of the construction by accumulation, the time samples at the output of the construction and accumulation module 1051 correspond to J consecutive samples of the time signal S' (t) obtained at the output of the transmission device over the duration of the OFDM block.

From the start of the OFDM block, a vector S' n-1[ ] of time samples is thus built up stepwise over each clock and accumulation period. The detection and comparison module 1052 then uses the vector S' n-1[ ] to detect the occurrence of peaks on the vector samples. The detection and comparison module 1052 deduces from the correction control information item POLn corresponding to the information item representing the polarity of the detected peak value by comparison with the time sample having the peak value among the samples of the time response of the current carrier n.

The correction module 1053 determines the correction dn of the constellation from the correction control information item POLn, as explained below with reference to the examples of fig. 5A to 5C.

The control information item POLn calculated from S 'n-1[ ] at index n contains the delay clock period represented by the black vertical bar 1054, which is associated with the symbol Xn, which is to be corrected to X' n by the control information item POLn and accumulated at the next clock pulse. In practice, the counting recorded at the frequency Fe is implemented at the output of the build and accumulate module 1051. Thus, at each edge of the clock, a new input (i.e. the corrected constellation symbol X 'n) is loaded and a new output is updated and corresponds to the time samples of the carrier n mapped by the corrected constellation symbol X' n accumulated respectively with the samples accumulated up to the previous order. During the clock cycle, the corresponding value of the time sample is then retained at the output.

Based on the complex control information item POLn, the correction module 1053 applies constellation correction, modifying the coordinates of the constellation symbol Xn by performing a complex displacement dn thereon.

In the next accumulation step, this complex displacement dn has the effect of reducing the detected peak amplitude relative to the case without correction. The accumulation will thus produce new values S' n of the series of parallel pre-constructed time samples by the regression constraint on the highest-ranked samples.

The method according to the invention is then repeated on each new constellation symbol. As the carrier is modified, a new peak to be modified will appear and be modified. The method according to the invention terminates at the end of the current OFDM block once all its constituent carriers have been scanned.

The build and accumulate module 1051 then re-initializes to process the next OFDM block.

5.2 detailed description and implementation of the different steps of the transmission method according to the invention

Fig. 2 shows specific steps implemented according to an embodiment of the present invention in order to generate a modification of the constellation of each carrier of a modulated OFDM block in order to reduce the PAPR of the transmitted signal, while fig. 3 shows an exemplary physical implementation of these steps.

Both of these aspects will be described in detail below.

5.2.1 description of the different steps of the method according to the invention.

For each carrier modulated by a constellation symbol, the method according to the invention implements processing operations equivalent to successively comprising at least the sub-steps represented in connection with fig. 4A to 4B.

In practice, PAPR reduction must be applied to the time signal that is "about to" be transmitted when there is no correction at the output of the OFDM transmission device.

From a series of constellation symbols modulating a carrier, the method according to the invention pre-constructs a real digital signal representative of a radio-frequency analog signal to be obtained at the output of a transmission device, each constellation symbol Xn being defined by a pair of values 400 (An, bn), said values 400 defining the coordinates of the constellation symbol Xn in the complex plane, such that Xn = An + j.bn.

To this end, according to the method of the invention, it is first considered to perform a frequency-time transformation (for the sake of illustration, consider An Inverse Fast Fourier Transform (IFFT) (41)) on the real (An) and imaginary (Bn) parts of each constellation symbol Xn, respectively modulating each of the N carriers of the OFDM block under consideration (N being An integer such that 0 ≦ N ≦ N-1). In practice, it is considered that M = N, and some of the N carriers may be reserved.

A representation 410 of the signal is then obtained in FIG. 4B, and the following expression at the IFFT output at T = k.Te, where 0 ≦ K ≦ K = N and K.Te = T (duration of the OFDM block under consideration):

then, shifting the real and imaginary parts by the term (42) to baseband is accomplished according to the equation and is represented 420 in FIG. 4B:

next, oversampling (43) of the real and imaginary parts in baseband by a factor L (e.g., L = 2) and a frequency equal to 2.Fe is performed according to the following equation:

where t = l.te/2 and 0 ≦ l < J = LN =2.n, which includes inserting 0 in one of every two samples, as shown in fig. 4B (430).

A low pass filter (44) of the real and imaginary parts is then implemented to retain only the baseband signal (440), according to the equation:

then, the modulation of each portion at a carrier frequency equal to Fe/2 is implemented (45) according to the following equation, as shown in fig. 4B (450):

it should be noted that, in general, the first and last carriers of an OFDM block are not modulated to avoid the problem of overlapping of the frequency spectrum associated with each OFDM block; thus only M carriers are actually mapped, where M ≦ N, M being an integer. It can also be considered that these carriers form part of a so-called reserved carrier to which no correction is applied, and in this case, more simply, M = N.

Signals modulated at a carrier frequency equal to Fe/2Thus equal to the inverse fast fourier transformed output signal oversampled at a frequency equal to 2.Fe and shown in fig. 4B (460).

Finally obtaining the expressionTo include within the factor:

wherein l is more than or equal to 0<2.N。

This equation, referred to hereinafter as the "equation of the junction", represents, for the OFDM block under consideration, the time signal that is "to be" transmitted without correction at the output of the OFDM transmission device.

This expression shows the time-symmetric behavior (Hermitian symmetry) illustrated by the following expression, where 0. Ltoreq. P.ltoreq.N:

thus, the 2.n series of samples for the cosine part of the N carriers are even, while the 2.n series of samples for the sine part of the N carriers are odd.

In addition to this, the present invention is,also shows a frequency symmetry characteristic described by the following expression, wherein 0. Ltoreq. Q. Ltoreq.N/2:

the result is that the first half series from the N carriers (for the first half series for N carriers) can be obtained by implementing the same operations with respect to exploiting temporal symmetry and inserting one complex time sample in every two samples) The portions of the latter half series (for 0. Ltoreq. Q. Ltoreq.N/2) are deduced.

Thus, the expression of the uncorrected junction signalIs subsequently used as a reference and updated by correcting the original values (An, bn) of the real and imaginary parts carrier by means of new corrected values (a 'n, B' n).

Obviously, the complex signal can be represented by its real and imaginary parts in the form of An + jBn or in the form ofIs expressed in terms of.

In fact, as indicated previously, the object of the present invention is to implement a "pre-constructed" (i.e. obtained) digital signal representative of the analog signal at the output of the transmission device from dynamic and peak angles, in other words to obtain a "picture" of the analog signal at the output of the transmission device, and to gradually modify, carrier by carrier, each constellation symbol modulating the carrier, so as to obtain a pre-constructed and modified signal with reduced PAPR.

According to the embodiment of the invention illustrated in fig. 2, a transmission method 20, for a current carrier with index N (N being an integer such that 0 ≦ N-1), making it possible to apply a modification of the modulation constellation to reduce the PAPR of the transmitted signal, comprises a first step of generating (21) J = l.n complex time samples associated with said current carrier with index N, where L is an integer and L ≧ 1.

Then, for a current carrier with index N (N being an integer such that 0 ≦ N ≦ N-1), the transmission method 20 according to the present invention comprises a selection from the group E that has been previously pre-constructed _n A second step of detecting (22) a maximum power peak. What is needed isThe above-mentioned group E _n Corresponding to the L.N pre-constructed real time samples, the time samples are obtained from the accumulation of L.N prior pre-constructed real time samples associated with n-1 carriers that have been previously modified when n ≧ 1, and the accumulation of L.N pre-constructed real time samples associated with the currently unmodified carrier when n = 0.

In the case of n =0, the first carrier of the OFDM block is concerned, so there is no set of l.n pre-constructed real time samples associated with a set of previously modified carriers. Thus, a default set E of l.n pre-constructed real time samples _D For peak detection, such that E0= ED.

For example, there may be a set of l.n pre-constructed real time samples of 0, in which case no peak will be detected and the correction information will be 0.

According to another variant, it can be considered that said default set E of l.n pre-constructed real time samples _D Corresponding to all the pre-constructed real time samples associated with the unmodified current carrier with index 0.

According to another variant, it can be considered that said default set E of l.n pre-constructed real time samples _D Corresponding to a predetermined group obtained, for example, according to a statistical model.

In the case of n ≧ 1, e.g., n =5, at L.N sets E of pre-constructed real time samples ₅ The time samples resulting from the accumulation of all l.n pre-constructed real time samples associated with the previously modified first 4 carriers (i.e., carriers with indices of 1, 2, 3, and 4) and the accumulation of l.n pre-constructed real time samples associated with the unmodified carrier with index of 0 are implemented in the detection step (22).

The method includes a step 220 of comparing between the detected peak and the time sample associated with the peak of the l.n complex time samples associated with the current carrier of index n provided by the generating step 21. The comparison step transmits a plurality of corrective control information items (Poln).

For example, the method is in a default set E of pre-constructed real time samples _D Detect a time corresponding to a fourth real numberThe peak value of the sample. The method will then pre-construct a default set E of real time samples _D Is compared (220) to a fourth complex time sample of complex time samples associated with the carrier index 0 provided by the generating step 21.

The comparison step 220 delivers a plurality of modified control information items POLn belonging to at least one of the following categories:

the real (respectively imaginary) part of the complex modification control information item, which is positive when the sign of the peak value, i.e. the sign of its complex coordinates, and the sign of the real (respectively imaginary) part of the corresponding complex time sample are the same,

the real (respectively imaginary) part of the complex modification control information item, which is negative when the sign of the peak value and the sign of the real (respectively imaginary) part of the corresponding complex time sample are opposite,

-the real (respectively imaginary) part of the complex modification control information item, which is 0 when the power amplitude of the time samples for which a peak has been detected is less than a predetermined threshold, or the current carrier with index n is said to be "reserved" and not to be modified.

In practice, some carriers, e.g. pilots or unmapped carriers at the edges of the spectrum, may not be modified to keep the mapped value associated with a constant or 0, respectively, unchanged. This applies in particular to pilots when the insertion is made before the PAPR reduction method according to the invention.

Then, for a current carrier of index N modulated by a constellation symbol Xn (N being an integer such that 0 ≦ N ≦ N-1), the transmission method 20 according to the present invention comprises a third step of modifying (23) the constellation symbol Xn of said current carrier of modulation index N as a function of the complex modification control information item. The modifying step conveys a modified current carrier of index n associated with a set of l.n modified complex time samples.

More precisely, the control information item makes it possible to select, from the three displacement possibilities for each real or imaginary part of the complex displacement, the complex displacement of the constellation symbol on the X and Y axes of the complex plane of the symbol constellation. In practice, when the complex modification control information item contains a positive real (respectively imaginary) part, the complex displacement effects a Negative (NEG) real (respectively imaginary) displacement. When the correction control information item contains a negative real (respectively imaginary) part, the complex displacement effects a Positive (POS) real (respectively imaginary) displacement. When the correction control information item contains a real (respectively imaginary) part (which means no correction), the complex displacement is realized as a (ZERO) real (respectively imaginary) displacement of 0.

The complex correction shift according to its two coordinates is thus achieved in the opposite direction of the PAPR peak formation.

Once the complex shift is selected as a function of the complex modification control information item, a summation of the coordinates (An, bn) of the constellation symbol Xn and the coordinates (dAn, dBn) of the selected complex shift dn is achieved. The correction step transmits new coordinates (a ' n, B ' n) of the corresponding corrected constellation symbol X ' n.

Furthermore, in accordance with an embodiment of the invention, in order to limit the complexity of the implementation, the absolute value of the complex displacement is set from one carrier to another of the OFDM block and corresponds to a predetermined value, for example a value lying between 0.25 and 0.5 times the distance between a constellation point and the boundary of the decision region.

Then, for a current carrier with index N (N being an integer such that 0 ≦ N ≦ N-1), the transmission method 20 according to the present invention comprises a fourth step of constructing (24) the L.N pre-constructed real time samples associated with the modified current carrier with index N modulated by X' N.

Then, for a current carrier with index N (N being an integer such that 0 ≦ N ≦ N-1), the transmission method 20 according to the present invention comprises a fifth step of accumulating (25). The accumulating step associates two by two l.n pre-constructed real time samples associated with a modified current carrier with index n with a set of l.n pre-constructed real time samples obtained from the accumulation of l.n pre-constructed real time samples associated with the previous n-1 carriers that have been previously modified when n ≧ 1 and the accumulation of l.n pre-constructed real time samples associated with the unmodified carrier when n = 0.

5.2.2 physical implementation of the different steps of the method according to the invention.

The values (An, bn) define constellation symbols Xn (Xn = An + j.bn). Compared to a conventional chain generating a signal S (t) from a series of values (An, bn) transformed by OFDM modulation (IFFT) as represented at the top of fig. 3, the method according to the invention generates correction values (a ' n, B ' n) which, after OFDM modulation (IFFT), result in a signal S ' (t) in which the PAPR peaks affecting S (t) have been reduced.

To this end, as described previously, the method according to the invention pre-constructs the complex time samples of the signal S' (l.te/2) by accumulating and simultaneously stepping over the duration of the OFDM block, as defined according to the "junction" equation mentioned previously. In other words, an uncorrected junction signal according to the previously mentioned "junction" equationThe expression (a) of (B) is then used as a reference and updated by modifying the original values (An, bn) of the real and imaginary parts carrier by means of new modification values (a 'n, B' n).

During said "pre-construction", the algorithm eliminates the signal peaks as they are formed by detecting them and then determining the direction of the correction to the constellation of symbols Xn +1 arriving in the opposite direction to the formation of these peaks.

According to the schematic diagram of fig. 3, an implementation of the method according to an embodiment of the invention comprises 6 processing modules, from 301 to 306, as described below.

More precisely, from the two portions (or coordinates) An and Bn of each OFDM block, correction module 301 receives as input the values of the regular constellation coordinates continuously and in synchronism with the clock (i.e. without correction from the modulation constellation previously applied, for example according to QAM modulation). The correction module determines three complex displacement possibilities (POS, ZERO, NEG) for each of the two portions that would have the effect of modifying the coordinates of the current constellation symbol Xn.

The method according to the invention thus makes it possible to select a complex shift scheme pair (dAn, dBn) to correct the constellation symbol Xn in three possibilities. By adding the original parts An and Bn, a new pair (a 'n, B' n) is obtained after the modification 23. As this new pair, applied in conjunction with the input of the IFFT (OFDM modulation 106) of fig. 1A, participates in the generation of signal S' (t). This new pair will also be reintroduced into the algorithm to update again the series of J = l.n =2.n pre-constructed real time samples of the current pre-constructed real time response (N = FFT size = IFFT size), which will be obtained after the IFFT and modulation implemented by the building block 306 and the accumulating block 305, in accordance with the previously mentioned "resultant" equation.

As described above, an operation is implemented between L.N pre-constructed real time samples and all complex time samples, the real time samples resulting from the accumulation of L.N pre-constructed real time samples associated with the previous n-1 carriers, which were previously corrected when n ≧ 1, and the accumulation of L.N pre-constructed real time samples associated with the unmodified carrier when n = 0; the complex time samples correspond to the cosine and sine portions of the current carrier under consideration with index n, which are stored in ROM memory or are algorithmically calculated by the module 302 for generating a multi-carrier signal. For each constellation symbol Xn considered, the module 302 for generating a multicarrier signal thus transmits a series of J = l.n =2.n cosine and sine complex time samples of the carrier N associated therewith in an OFDM block of size N. The module 302 thus constructs 21 l.n complex time samples per carrier n.

At each clock pulse, all complex time samples of the signal S' (l.te/2) are thus stepwise pre-constructed by accumulating 25 the current result in the accumulation module 305 according to the above equation, unlike the result of the correction operation previously implemented.

For example, the complex time samples corresponding to the cosine and sine portions of the carrier with index n constitute a vector COS as defined below _n []And SIN _n []The elements (c):

the signal vector S' n [ ] is thus obtained from the accumulation 25 by the accumulation module 305 of l.n =2.n real time samples constructed in module 306 associated with the previous n-1 carriers previously modified when n ≧ 1, and 2.n constructed real time samples associated with the unmodified carrier when n = 0. The signal vector S' n [ ] is then used as input to block 304 to detect the peak PEAKn, which is expressed below for index n:

S _n ′[]<＝S′ _n-1 []+A′ _n-1 .COS _n-1 []-B′ _n-1 .SIN _n-1 [] 0<n<N

wherein, for example, S' ₀ []＝0。

Because the accumulation module 305 has recorded, the operator "< = in the above equation means that the term corresponding to the left side of the output of the accumulation module 305 will meet the requirement of the term corresponding to the right side one clock cycle later.

Thus, when the constellation symbol series has reached symbol Xn, indexed n, for index n, the median value S 'n [ ] of the vector of J = l.n =2.n pre-constructed real time samples, representing S' (l.te/2), is obtained at the output of the recorded accumulation module 305.

According to an embodiment, S 'when N = N-1' _N-1 []The real time samples containing S' (l.te/2) constructed for the N carriers of the OFDM block under consideration, except for the last value discarded, and the accumulation module 305 would reset to 0 for the new OFDM block, for example.

At each clock pulse Clk representing the processing rate from one carrier to another, J = l.n =2.n results are loaded into accumulation module 305 in accordance with the "junction" equation mentioned earlier. In parallel, the complex time samples of the cosine and sine portions provided by the module 302 for generating a multicarrier signal, previously at index n-1, are replaced by time samples associated with a carrier of index n to be subsequently modulated by a constellation symbol X ' n, which is a pair of Xn characterized by new corrected coordinate values (a ' n, B ' n).

By searching for the real time sample of maximum amplitude among the 2.N pre-constructed real time samples of the signal vector S' n [ ], the module 304 for detecting a peak value PEAKn makes it possible to detect (22) a signal peak that may occur from the output of the accumulation block 305, the time samples resulting from the accumulation of all 2.N pre-constructed real time samples associated with the previous n-1 carriers that have been previously corrected when n ≧ 1, and 2.N real time samples associated with the uncorrected carrier when n = 0. The real time samples are determined by the following equation:

PEAK _n ＝max _0≤l<2.N (S _n ′(l))。

the detection module 304 then provides information related to the detected peak to the comparison module 303 which will determine 220 whether the cosine and sine complex time samples of the current carrier corresponding to the peak, indexed n, have a complex polarity POLn = POLAn + j. This polarity is then used as a complex correction control information item. For example, POLn is defined as:

if the polarity is the same between the cosine (sine) part of the complex time samples and the detected peak, POLAN (POLBN) is positive and correction by a Negative (NEG) complex shift dAn (dBn) is chosen. Conversely, if the polarity is opposite, polar (POLBn) is negative and correction is selected by a positive complex shift da n (dBn) to obtain corrected portions a 'n and B' n of constellation symbol Xn.

In case the amplitude of the cosine (or sine) part of the complex time sample, which coincides with the detected peak, is too small with respect to a certain threshold Pseuil (threshold amplitude corresponding to POLn) and will not be able to provide a significant peak reduction, the ZERO real (respectively imaginary) displacement of the complex displacement will be applied, i.e. no correction will be made.

If the correction is not sufficiently effective, preferably no modification is made to the constellation symbol under consideration. In fact, in this case, the modification degrades the bit error rate or meaningless increases the average power by complex shifts in the original constellation.

In contrast, in the case of the TR technique described in connection with the prior art, the threshold Pseuil is not necessary, since the pilot will not add any interference to the useful signal, whatever the correction benefit will be obtained.

The peak detection module 304 thus provides the location of the real time samples, which shows the peak level and its sign. For the An (Bn) portion, the following table summarizes the direction of the complex shift, which will be referred to as COS _n (l _peak ) (accordingly (-SIN) _n (l _peak ) ) to provide a value Pseuil corresponding to a threshold amplitude of POLn in practice between 0.15 and 0.3:

if the goal is to move the constellation symbols and remain in the original constellation, a positive real displacement dAn (imaginary displacement dBn) for An (Bn) is reflected as a real displacement to the right of the constellation point on the complex plane (corresponding to An upward imaginary displacement), whereas a negative real (imaginary) displacement is to the left (downward).

Conversely, if the goal is to move a constellation symbol out of the original constellation as in the TI-CES technique, a point is moved to the opposite point of the extended constellation only if the sign of the An (Bn) portion is opposite to the sign of the direction of the real displacement da (imaginary displacement dBn) (positive or negative is chosen as a function of the polarity polar (POLBn)); otherwise, the point is not modified.

5.2.3 results of the method according to the invention

80% to 90% of the signal peaks that result in strong PAPR can be processed with the previously described method.

It should be noted, however, that some signal peaks occur very late in the formation of the OFDM block, which can only be partially corrected by the method according to the invention, when the latter has a rather limited capability of correction on a carrier-by-carrier basis.

Therefore, in order to obtain an effective correction, the PAPR peaks of the signal must appear early enough in the "pre-construction" process to be effectively corrected later, or it is necessary to shift the constellation symbols out of the original constellation as in the CES method.

Conversely, when processing the first carrier, very few high peak levels can be discerned and the correction capability of the method is underutilized.

Furthermore, it should be noted that the method according to the invention can be combined with the rather simple "clipping" or limiting techniques described above in connection with the prior art to improve the reduction of PAPR.

5.3 example description of a New constellation obtained according to the invention

For the TI-CES, CD, ACE and TR techniques in the prior art, the method according to the invention modifies the modulation constellation of the carrier in the frequency domain before IFFT to the result to obtain a signal with reduced PAPR in the time domain.

An advantage of the method according to the invention is the flexibility of constellation modification. In fact, any type of correction can be applied according to the invention provided that it can be reflected in a controlled complex displacement of the real and/or imaginary part of the constellation symbols associated with the carrier.

Thus, displacements in the positive or negative directions on the X and Y axes of the complex plane may cause constellation points to remain within their decision regions or in the original constellation, or to be shifted out.

In connection with fig. 5A, the method according to the invention makes it possible to apply complex displacements equally both inside (51) and outside (52) the original constellation.

Thus, implementation of the method according to the invention delivers two new more general classes of constellation modifications:

class 51ICS (intra constellation transfer), where the moved constellation points remain inscribed in the original constellation when the applied modification remains moderate, this class including, in particular, the prior art CD technique described previously, and

class 52OCS (outer constellation transfer), where points are moved outside the original constellation, this class comprising TI-CES and ACE techniques in the prior art as described previously

For the TR technique (according to the prior art described above), there is no concept of constellation, and the constellation of the peak reduction Preamble (PRT) carrier is defined only to reduce PAPR, which can also be controlled by the method according to the present invention.

Constellation modification implemented in accordance with the present invention can thus replace any of the previously described prior PAPR reduction techniques, regardless of whether all of the advantages seen by separating each of them are taken therein.

One advantage of the proposed system is that it is possible to obtain more benefits by making it possible to associate several technologies, the different drawbacks of which can be neutralized to some extent.

By way of example, fig. 5B shows that all complex shifts are allowed for QPSK (53), MAQ16 (54), MAQ64 (55) constellations resulting from the original combination of the so-called ICS constellation modifications (small arrows inward) inscribed in the modified original constellation retained by the shifted constellation points and the so-called OCS constellation modifications (large arrows outward) shifted out of the original constellation.

Similarly, fig. 5C shows two constellations MAQ16 (56) and MAQ64 (57), which contain ICS and OCS complex shifts. For conventional points affected by complex displacements outside the original constellation, there may be two opposite correction schemes with respect to one of the real or imaginary axes, an original white point and an outward "spreading" black point.

The number and location of the new "extended" black points are determined in such a way that the increase in average power of the constellation remains limited (the constellation is inscribed at most in a circle), but that good correction capability is maintained in each case of MAQs 16 (56) and 64 (57) by moving a fraction of the points, which is about one-third or one-half of the points moved into the original constellation, out of the original constellation.

The method according to the invention thus makes it possible to obtain two new constellation types, respectively called "ICS" or "OCS", depending on whether a constellation point is shifted in or out from the original constellation, which combines the advantages of the different prior art.

5.4 description of a Transmission device according to the invention

A simplified structure of an apparatus for transmitting an OFDM signal representative of a source OFDM signal comprising OFDM blocks each comprising a set of N carriers for implementing the transmission technique according to the invention is shown in connection with fig. 6.

Such a transmission device comprises a memory module 60 comprising a buffer M, for example a processing unit 61 equipped with a microprocessor μ P and driven by a computer program 62 for implementing the transmission method according to the invention.

At initialization, the code instructions of the computer program 62 are loaded, for example, into a RAM memory M before being executed by the processor of the processing unit 61. The processing unit 61 receives as input complex symbols Xn onto which data representing the source signal is mapped. According to the instructions of the computer program 62, the microprocessor of the processing unit 61 implements the steps of the transmission method described previously to achieve a modification of the modulation constellation with the aim of reducing the PAPR of the transmission signal S' (t). To this end, the transmission apparatus comprises:

-a mapping module for mapping data representing the source signal onto complex symbols Xn belonging to a constellation, 0 ≦ n < M, n and M being integers,

-a transformation module for transforming the M symbols Xn into M modified symbols X 'n such that X' n = Xn + dn, dn being a complex modification,

-an OFDM modulator having N carriers for generating an OFDM signal from M modified symbols X' N mapped on the M carriers, wherein M ≦ N, N being an integer.

The transformation module includes:

-a construction and accumulation module for accumulating J time samples corresponding to J samples of an order n carrier mapped by X' n onto existing J samples, J being an integer, J >0,

-means for detecting a peak value on J samples at the output to the n-1 th order building and accumulating means and comparing this peak value with a simultaneous time sample of J time samples of the n-order carrier and delivering a complex correction control information item,

According to an embodiment, the build and accumulate module comprises:

-an accumulation module for accumulating the data received from the input module,

and the transform module further comprises:

-means for generating J complex time samples associated with the current carrier with index n, J being an integer.

These means are driven by the microprocessor of the processing unit 61.

Claims

1. A method (20) for transmitting an OFDM signal comprising OFDM blocks and each block comprising a set of N carriers, the processing of said OFDM signal prior to transmission comprising:

-transforming the M symbols Xn into M modified symbols X 'n such that X' n = Xn + dn, where dn is a complex modification,

mapping the M corrected symbols X' N onto M of the N carriers of the OFDM modulator to generate an OFDM signal, N being an integer, M ≦ N,

such that the transformation comprises:

-an initialization step of the accumulator,

and for each nth order carrier, where n is from 1 to M-1, comprising:

-a step (1051) of accumulating, in an accumulator, J time samples corresponding to J samples of the n order carriers mapped by X' n, respectively with the J samples already present,

-a step (1052) of detecting a peak value among J samples obtained in the accumulating step from 1 st order to n-1 st order and comparing the peak value with a simultaneous time sample among J time samples of the n th order carrier, the step (1052) transmitting a complex modification control information item (Poln),

-a correction step (1053) of determining a complex correction value dn as a function of the complex correction control information item (Poln) to obtain a corrected symbol X' n.

2. The transmission method (20) of claim 1, further comprising the step of generating J time samples (21) associated with the current carrier of order n, wherein the detecting and comparing step (1052) is carried out with a delay (1054) of a determined duration corresponding to a clock period in combination with the step of generating J time samples (21) associated with the current carrier of order n.

3. The transmission method (20) according to claim 1, wherein the detected peak is a power peak, the J time samples of the nth order carrier are complex numbers, and wherein the complex modification control information item (Poln) belongs to at least one of the following categories:

-the real, or respectively the imaginary part of the complex modification control information item is positive when the sign of the peak value is the same as the sign of the real, or respectively the imaginary part of the corresponding complex time sample,

-the real, or respectively the imaginary part of the complex modification control information item is negative when the sign of the peak value and the real, or respectively the sign of the imaginary part of the corresponding complex time sample are opposite,

the complex modification control information item has a real, or respectively imaginary, part of 0 when the power amplitude of the time sample of the detected peak is smaller than a predetermined threshold, or when the current carrier with index n is said to be "reserved" and not to be modified.

4. The transmission method (20) according to claim 1, wherein the step of modifying (1053) implements a summation of the coordinates of a constellation symbol Xn and the coordinates of a complex displacement of said constellation symbol on the X and Y axes of the complex plane of the constellation of the symbol, said complex displacement being selected by means of an item of complex modification control information, from complex displacements belonging to at least one of the following categories:

-when the complex modification control information item contains a positive real number, or respectively an imaginary part, the real number, or respectively the imaginary displacement, of the complex displacement is negative;

-when the complex modification control information item contains a negative real number, respectively an imaginary part, the real number, respectively the imaginary displacement, of the complex displacement is positive;

-when the complex modification control information item contains a real number, respectively an imaginary part, which is 0, the real number, respectively the imaginary displacement, of the complex displacement is 0.

5. Transmission method (20) according to claim 4, characterized in that the absolute value of the real, or respectively the imaginary part, of the complex displacement is determined from one carrier to another carrier of the OFDM block and corresponds to a predetermined value.

6. The transmission method (20) according to claim 1, wherein, for a given carrier n mapped by X' n, the transform implements a discrete inverse fourier transform computed carrier by computing J time samples.

7. Transmission method (20) according to claim 1, wherein the number of time samples J is derived from an oversampling resulting in J = NL samples by an integer factor L ≧ 1.

8. The transmission method (20) of claim 7, wherein the transmission method implements the following equation for L =2:

wherein:

-A _n and B _n Is the real and imaginary parts of the constellation symbol that modulates the current carrier with index n,

-S is all pre-constructed real time samples associated with an OFDM block,

fe is the sampling frequency of the data representing the source signal, and

-0≤l<N.L＝2.N。

9. an apparatus (1000) for transmitting an OFDM signal, the apparatus comprising:

-a mapping module for mapping (103) data representing the source signal onto complex symbols Xn belonging to a constellation, 0 ≦ n < M, n and M being integers,

-a transformation module for transforming (105) the M symbols Xn into M modified symbols X 'n such that X' n = Xn + dn, where dn is a complex modification,

-an OFDM modulator having N carriers for generating (106) an OFDM signal from M corrected symbols X' N mapped on the N carriers, N being an integer, M ≦ N,

the transformation module includes:

-a build and accumulate module (305, 306) for accumulating J time samples corresponding to J samples of an order n carrier mapped by X' n onto the existing J samples,

-means (304, 303) for detecting the peak of J samples at the output of the building and accumulation means from order 1 to order n-1 and comparing this peak with the simultaneous time samples of J time samples of the carrier of order n, which conveys a complex item of modification control information (Poln),

-a correction module (301) for determining a complex correction value dn as a function of the complex correction control information item (Poln) to obtain a corrected symbol X' n.

10. The apparatus (1000) for transmitting OFDM signals according to claim 9, such that said building and accumulating module (305, 306) comprises:

-a module (306) for constructing J time samples of the X' n mapped current carrier of order n, and

-an accumulation module (305),

and such that said transformation module further comprises:

-means (302) for generating J time samples associated with a current carrier of index n.

11. An information medium containing program instructions adapted to perform the following steps when loaded and executed in an OFDM transmission apparatus:

-transforming the M symbols Xn into M modified symbols X 'n, such that X' n = Xn + dn, dn being a complex modification,

mapping the M corrected symbols X' N onto N carriers of an OFDM modulator to generate an OFDM signal, where N is an integer, M ≦ N,

such that the transformation comprises:

-an initialization step of the accumulator,

and for each nth order carrier (n from 1 to M-1) comprises:

-a step (25) of accumulating, in an accumulator, J time samples corresponding to J samples of the n order carriers mapped by X' n, respectively with the J samples already present,

-a step (22) of detecting a peak value among the J samples obtained in the accumulation step from order 1 to order n-1 and comparing this peak value with a simultaneous time sample among the J time samples of the carrier of order n, this step (22) transmitting a complex modification control information item (Poln),

-a correction step (23) of determining a complex correction value dn as a function of the complex correction control information item (Poln) to obtain a corrected symbol X' n.

12. A method for multi-carrier transmission (20) of an OFDM signal representing a source OFDM signal comprising OFDM blocks and each block comprising a set of N carriers, characterized in that for a current carrier modulated by a constellation symbol Xn with an index N, where N is an integer such that 0 ≦ N-1, the method comprises the following steps performed in the frequency domain before performing an Inverse Fast Fourier Transform (IFFT):

-constructing (21) a set of M complex time samples associated with the current carrier with index n, M being an integer,

-detecting (22) a maximum power peak from a previously stored set of M pre-constructed real time samples, said time samples resulting from the accumulation of all M pre-constructed real time samples associated with the previously modified previous n-1 carriers when n ≧ 1 and the accumulation of M pre-constructed real time samples associated with the unmodified carrier when n =0, and transmitting a complex modification control information item (Poln) taking into account said set of M complex time samples associated with said current carrier with index n,

-modifying (23) the constellation symbol Xn modulating the current carrier with index n as a function of the complex modification control information item and transmitting a modified current carrier with index n associated with a set of M modified complex time samples,

-constructing (24) a set of M pre-constructed real time samples associated with the modified current carrier with index n from the previous set of M modified complex time samples,

-an accumulation (25) of said M pre-constructed real time samples associated with said modified current carrier with index n, two by two with said previously stored set of M pre-constructed real time samples, obtained from the accumulation of all M pre-constructed real time samples associated with the previous n-1 carriers modified when n ≧ 1 and the accumulation of M pre-constructed real time samples associated with the unmodified carrier when n =0, and

-storing (26) the result of said accumulation.

13. Transmission method according to claim 12, characterized in that it comprises, for an OFDM block, a preliminary step of initialising 0 the storage module to be used for said step of storing the result of the accumulation applied to each of the N carriers of said OFDM block.

14. Transmission method according to claim 12, characterized in that said detection step (22) of the transmission of a complex item of modification control information (Poln) is carried out in conjunction with said step (21) of constructing a set of M complex time samples associated to said current carrier with index n, with a delay (1054) corresponding to a clock period of a predetermined duration, where M is an integer.

15. The transmission method according to claim 12, wherein the complex modification control information items are obtained by comparing (220) the detected maximum power peak with corresponding complex time samples of the set of M complex time samples associated with the current carrier with index n, the complex modification control information items belonging to at least one of the following categories:

-the real, respectively imaginary part of the complex modification control information item is positive when the sign of the peak value is the same as the sign of the real, respectively imaginary part of the corresponding complex time sample,

-the real, respectively imaginary part of the complex modification control information item is negative when the sign of the peak value and the real, respectively imaginary part of the corresponding complex time sample are opposite,

-the complex modification control information item has a real, or respectively imaginary, part of 0 when the power amplitude of the detected peak's time sample is less than a predetermined threshold, or said current carrier with index n is said to be "reserved" without modification.

16. Transmission method according to claim 15, characterized in that said correction step (23) implements a summation of the coordinates of said constellation symbols with coordinates representing complex displacements of said constellation symbols on the X and Y axes of a complex plane of the constellation of said symbols, said complex displacements being selected by means of said complex correction control information items from complex displacements belonging to at least one of the following categories:

-when the complex modification control information item contains a positive real number, or respectively an imaginary part, the real number, respectively the imaginary displacement, of the complex displacement is negative;

-when the complex modification control information item contains a negative real, respectively imaginary part, the real, respectively imaginary displacement of the complex displacement is positive;

-when the complex modification control information item contains a real, respectively imaginary part of 0, the real, respectively imaginary displacement of the complex displacement is 0.

17. The transmission method according to claim 16, characterized in that the absolute value of the real, respectively imaginary part of the complex displacement is determined from one carrier to another carrier of the OFDM block and corresponds to a predetermined value.

18. Transmission method according to claim 12, characterized in that it successively carries out at least the following sub-steps:

-applying (41) an inverse fast Fourier transform to the real and imaginary components of the constellation symbols, centered at a frequency equal to Fe/2,

-transposing (42) the real and imaginary components to baseband,

-oversampling (43) the real and imaginary components in baseband at a frequency equal to 2.Fe,

-low-pass filtering (44) of the real and imaginary components,

-modulating (45) said components at a carrier frequency of Fe/2.

19. The transmission method of claim 12, wherein the transmission method implements the following equation:

wherein:

-A _n and B _n Is a star modulating a current carrier with index nThe real and imaginary components of the anchor symbols,

-S is all the pre-constructed complex time samples associated to an OFDM block,

fe is the sampling frequency of the data representing the source signal, and

-0≤l≤M＝2.N。

20. an apparatus for multicarrier transmission of an OFDM signal representing a source OFDM signal, said source OFDM signal comprising OFDM blocks and each block comprising a set of N carriers, characterized in that for a current carrier modulated by a constellation symbol Xn with an index N, where N is an integer such that 0 ≦ N ≦ N-1, the apparatus comprises the following modules implemented in the frequency domain before implementing an Inverse Fast Fourier Transform (IFFT):

-means (302) for constructing a set of M complex time samples associated with the current carrier with index n, M being an integer,

-means (304) for detecting the maximum power peak in a previously stored set of M pre-constructed real time samples, said time samples resulting from the accumulation of all M pre-constructed real time samples associated with the previously modified previous n-1 carriers when n ≧ 1 and the accumulation of M pre-constructed real time samples associated with the unmodified carrier when n =0, and transmitting a complex modification control information item (Poln) taking into account said set of M complex time samples associated with said current carrier with index n,

-a module (301) for modifying the constellation symbols Xn modulating the current carrier with index n as a function of the complex modification control information item, which module (301) transmits a modified current carrier with index n associated with a set of M modified complex time samples,

-a module (306) for constructing a set of M pre-constructed real time samples associated with a modified current carrier of index n from a previous set of M modified complex time samples,

-an accumulation module (305) for associating, two by two, said M pre-constructed real time samples associated with said modified current carrier with index n with said previously stored set of M pre-constructed real time samples resulting from the accumulation of all M pre-constructed real time samples associated with the previous n-1 carriers modified when n ≧ 1 and the accumulation of M pre-constructed real time samples associated with the unmodified carrier when n =0, and

-a module (60) for storing said accumulation result.